Re: [PATCH v13 1/4] :mm/vmalloc: allow to set node and align in vrealloc

[Vitaly Wool <vitaly.wool@xxxxxxxxxxx>]

> On Jul 15, 2025, at 4:37 PM, Matthew Wilcox <willy@xxxxxxxxxxxxx> wrote:
> 
> On Tue, Jul 15, 2025 at 03:57:24PM +0200, Vitaly Wool wrote:
>> +void *__must_check vrealloc_node_align_noprof(const void *p, size_t size,
>> + unsigned long align, gfp_t flags, int nid) __realloc_size(2);
>> +#define vrealloc_node_noprof(_p, _s, _f, _nid) \
>> + vrealloc_node_align_noprof(_p, _s, 1, _f, _nid)
>> +#define vrealloc_noprof(_p, _s, _f) \
>> + vrealloc_node_align_noprof(_p, _s, 1, _f, NUMA_NO_NODE)
>> +#define vrealloc_node_align(...) alloc_hooks(vrealloc_node_align_noprof(__VA_ARGS__))
>> +#define vrealloc_node(...) alloc_hooks(vrealloc_node_noprof(__VA_ARGS__))
>> +#define vrealloc(...) alloc_hooks(vrealloc_noprof(__VA_ARGS__))
> 
> I think we can simplify all of this.
> 
> void *__must_check vrealloc_noprof(const void *p, size_t size,
> unsigned long align, gfp_t flags, int nid) __realloc_size(2);
> #define vrealloc_node_align(...) \
> alloc_hooks(vrealloc_noprof(__VA_ARGS__))
> #define vrealloc_node(p, s, f, nid) \
> alloc_hooks(vrealloc_noprof(p, s, 1, f, nid))
> #define vrealloc(p, s, f) \
> alloc_hooks(vrealloc_noprof(p, s, 1, f, NUMA_NO_NODE))
> 
> 

In this case, to keep things buildable an each step we will need to modify slub.c in this patch. Since we change slub.c in the next patch in the series I would suggest that we keep things simple (== as they are now, even if it means some redundant macros have to stay). I can come up with a macro simplification like yours when this series is accepted.

Thanks,
Vitaly